8!) Mr, Nixon on the Theory of the '{Au(j. 



It is to be remarked that any one or more sections of a cylin- 

 drical column of air, uniformly of a given temperature, may be 

 heated or cooled without disturbing in the least the pressure of 

 the strata incumbent on that section, or of those on which it 

 rests.* That the density of the lower strata of the atmosphere 

 is occasionally inferior (especially in the middle of a cloudless 

 day declining rapidly in temperature towards morning and 

 evening), to that of the strata immediately superincumbent is 

 proved by the terrestrial refractions being frequently negative 

 When the intercepted arc is inconsiderable. 

 ' If we admit the progression of the diminution of temperature 

 as we ascend to be geometrical, and of such a nature that the 

 decrement for the same number of feet is greater at the base 

 than at the summit, the column will evidently contain a greater 

 quantity of air inferior than superior in temperature to the mean 

 of the detached thermometers at the extremes, and half their 

 som exceeding the true mean temperature, will introduce an 

 error in excess in the calculated altitude of the object. Still as 

 we have no experiments authorising us to conclude that the 

 diminution of temperature is in geometrical progression ; as we 

 are even ignorant whether the rate be greater at the base or 

 summit, and, what is more to the point, being well aware that 

 numerous local and other causes will disturb and render its 

 nature undistinguishable, we may be spared the trouble of cal- 

 culating any corrections, and content ourselves with considering 

 the mean of the thermometers at the extremes as the proper 

 temperature for calculation. f We cannot, however, too strongly 

 insist that the principal errors of barometrical measurements are 

 the result of an incorrect estimate of the mean temperature. It 

 has also been justly observed by Prof. Play fair, that when the 

 horizontal distance of the barometers is considerable, the tem- 

 perature of the air at the lower station may not accord with that 

 at the base of the column of air immediately under the upper 



• If wc increase the temperature of a fluid, confined in a truly cylindrical vessel, to 

 such a decree as to double the volume, the height will be increased^ and the density 

 diminished^ in the same ratio, and the pressure at the base wiU remain unaltered. But 

 if tlie figure of the vessel be that of tlie frustum of an inverted ' 

 pyramid, then will tlic increment of volume be insufficient, 

 on account of the increasing capacity of tiie vessel upwards, to 

 augment the depth of the fluid in the same ratio. The ab- 

 solute pressure being as the height multiplied by tlie density, 

 there is a consequent loss of pressure. Hence it must be ad- 

 mitted that an increase in any ratio of the volume of the 

 entire atmosphere (resulting from an elevation of its temper- 

 ature throughout its mass) will not cause, on account of its 

 RI>herical figure, an augmentation of its heiglit in the same 

 proportion ; and that the pressures at the base of an atmosphet-erunifona^^ aaoprding in 

 temperature with our climate would be less in summer and at noon, than in winter and 

 at night. This view of the subject, if correct, may serve to account for the horary 

 oscillations of the barometer. . * i 



f Dr. Horsley considers the method of taking the ineanof the temperature* J^t the 

 extremes as only an a])proximation, yet sufficiently exact.— (Phil. Trans, vol. Lxiv.) 



